NAD83 vs. WGS84

[+gcfishguy]

I was just wondering if there was an easy answer to the question, "just how 'off' will you be using different datums?"

 

I know the standard for geocaching is Lat/Lon hddd mm.mmm and WSG84, but I will be helping out staking some claims this summer, and the prospectors in N.B. are told to use UTM NAD83 for coordinates on claim posts when registering the claims.

 

Lat/Lon to UTM is just a conversion, I realize, but how different will things be if I start switching back and forth between WGS84 and NAD83?

 

I use my GPS a lot, for camping spots, fishing, boating, and of course, caching....and I also use TopoCanada and GSAK.

 

I just want a qualified answer on:

 

1. If he's set to NAD83 and I'm using WGS84, will we be off (compared to each other)?

2. If I start bouncing back and forth between the 2, will I run into issues where some of my waypoints and routes will end up in the wrong place in the software? (Because of using 2 different Datums)

 

As a test, I waypointed a spot in topocanada with the software set to WGS84, and then changed it to NAD83 and created another waypoint with the identical coordinates. Well, the waypoints were right on top of each other. (As I had hoped)

 

Now, was my test a valid test? I had have read on these forums that people were quite a ways off by using the wrong datum, but I never got to see what datum they were using.

 

Also, it could be that in a different area of the country, there may be a big difference between those 2 waypoints...but the same test gave the same result on the west coast.

 

Now, maybe it's TopoCanada just making some sort of a correction on my behalf, as an idiot-proof thing that's built into the software...

 

If my topocanada test is a valid one, then I'll just switch back and forth between UTM and Lat/Lon when I need to, and the datum should be irrelevant. (At least NAD83 vs. WGS84) Other combinations will surely put a person way out in the rhubarb....

 

Anyway, could someone confirm that there is no difference (for my intentions) between NAD83 and WGS84?

 

Thanks in advance!

 

<><

 

**Edit - I just went and did some quick reading and found that my memory isn't that good. I was sure I read (a while back) that the 2 could be up to 50m apart....but it looks like it's more like a meter.

Edited May 29, 2007 by gcfishguy

[Exit106]

I'd like to say I new the answer, but obviously someone really does know.

http://www.mentorsoftwareinc.com/resource/Nad83.htm

 

Pierre

[+gcfishguy]

Yes, it makes for some pile of reading, doesn't it?

 

I guess I've now answered my own question...unless you've shelled out tons of clams for a sub-meter unit, the difference isn't enough to care about...the error factor of the GPSr is more than the difference between WGS84 and NAD83. Plus, I also saw a few articles that say that consumer GPS manufacturers actually treat the 2 datums exactly the same....

 

I also found what I THOUGHT I remembered....turns out that it's NAD27 and WGS84 that can really throw you for a loop (as much as 200m) ...not NAD83.

 

So, I'll go hang my head in shame for thinking that I remembered something, rather than going and doing some reading. hehe

[+Kewaneh & Shark]

The difference between NAD83 and WGS84 is approximately one meter across North America. A bit more in some places, a bit less in others. For all intents and purposes, in the world of recreational GPSrs, the datums are virtually identical, as most recreational units only have +/-10 meter accuracy, +/-5 meters with WAAS activated. If your GPSr is set to NAD83 and you look for a cache with coordinates established in WGS84, your handheld will still get you within the targeted search radius. Unless you are using a survey-grade, sub-centimeter GPS system ($60,000+) the differences between NAD83 and WGS84 are next to irrelevant.

 

The differences between the NAD27 & NAD83 are very subtantial. Differences of 200-400 feet are common. In some areas up to 700-800 feet, and possibly more. NAD27 does not play well with Geocaching.

 

Hope this helps.

 

- Kewaneh

[Xlobsterman]

What is the relationship between the NAD83 and WGS84 coordinates?

 

One of the questions most frequently asked by GPS users is: How do I transform coordinates between the frames defined by NAD83 and WGS84?

In effect, the GPS-user is trying to convert the positional coordinates that the receiver outputs or the manufacture proprietary software gives from one particular frame into the other. First of all, one should understand that the 3-D Cartesian frames to which the coordinates of the NAD83 and WGS84 refer are not identical. Their origin, axes orientation in space, and the unit of scale differ. Why? Simply, because the definitions of these two frames are based on different sets of observations, processing algorithms, and perhaps, geodetic assumptions.

 

In practice, any terrestrial frame is defined by a set of points with assigned coordinates and velocities. These parameters materialize the reference frame in question. The velocities are important to change positional coordinates from one epoch to another. Thus, it is critical to understand that if we refine the coordinates and/or the velocities of the defining points by introducing new and more accurate observations, then, the origin, orientation and scale of the frame may change, but to a small degree. That is why sequence of reference frames has been developed. The latest version of NAD83 and WGS84 frames are specifically named NAD83 (CORS96) and WGS84 (G1150).

 

Thus, when the receiver tells us that the computed coordinates using GPS observations are referred to the WGS84, they actually mean WGS84(G1150). On the other hand, if we use CORS coordinates data, or the OPUS service, the NAD83 reference frame to which these coordinates refer is actually NAD83 (CORS96).

 

Consequently, rigorously speaking, a transformation between NAD83 and WGS84 should be interpreted as a transformation between NAD83 (CORS96) and WGS84 (G1150). Transformations between older realizations of NAD83 and WGS84 are now outdated and will not be discussed here.

The question is then reduced to: How do we transform between NAD83 (CORS96) and WGS84 (G1150)?

 

In absolute positioning, the frame WGS84 (G1150) is materialized not by the coordinates of marks on the surface of the Earth - as the NAD83 (CORS96) is - but by the coordinates of points in space, namely, the 3-D location of the GPS satellites (given by their precise or broadcast satellite ephemerides). Recent investigations [True, 2004] have shown that for all practical purpose, the WGS84 (G1150) frame is identical to the International Terrestrial Reference Frame of year 2000 (ITRF2000). Thus, except for very accurate geophysical calculations (crustal dynamics; tectonic strain determination, etc.), the WGS84 (G1150) is supposed to be aligned with the ITRF2000. In another words, both have the same origin, orientation, and scale.

 

Once the WGS84 (G1150) and ITRF2000 are assumed equivalent [ ≡ is the mathematical symbol for equivalent], the question can be taken one step further and finally be formulated as: How do we transform between NAD83 (CORS96) and ITRF2000?

 

In order to do this transformation, the epoch at which the coordinates defining the frames were determined must be known. The National Geodetic Survey (NGS) provides coordinates and velocities of all CORS stations on these two frames: ITRF2000, epoch 1997.0, and NAD83 (CORS96), epoch 2002.0, for the conterminous states. The epoch 2003.0 was selected for the state of Alaska due to the displacements originated by the Denali earthquake on November 2, 2002. The value of these coordinates and velocities as well as a description of previous ITRF frames can be consulted in the following CORS Web page: http://www.ngs.noaa.gov/CORS/metadata1/.

 

Let’s assume now that we want to transform coordinates between the WGS84 (G1150) ≡ ITRF2000 and NAD83 (CORS96). The first thing that we need to keep in mind is that, in theory, the WGS84 (G1150) or ITRF2000 coordinates are the first one computed by the receiver software. However, they are originally obtained at the time of the observation, that is, the epoch corresponding to the mid-point of the observation window during the GPS data were collected. Consequently, we cannot compare coordinates from a set of observations taken today to results we determined one year ago. These two sets of coordinates must be reduced to a common epoch before the comparison is made. Let’s assume that we want to compare the coordinates we obtained today referred to the ITRF2000 to the ones at epoch 1997.0. Some receivers may be already doing this transformation from epoch to epoch and the final result may be already expressed in the frame WGS84 (G1150) ≡ ITRF2000 at epoch 1997.0. This will simplify the work involved, otherwise we should know the velocity of the GPS station due to the rotation of the tectonic plate on which is located. These velocities are not rigorously known until we have constantly monitored the point for a number of years. This is not generally the case and we must rely on geophysical models. The most common model used these days for correcting for plate rotations is called NNR-NUVEL-1A. This model provides the parameters required to correct for plate motion for all major plates (macroplates) forming the crust of the earth.

 

Assume for an instant, that the velocities of the three components vx, vy, and vz, are known in cm/year for a particular point, then, the transformation between epochs could be simply written:

x (epoch 1997.0) = x (epoch today) + vx * ( 1997.0 - epoch today)

with similar equations for y and z.

 

Once we have all observed points referred to a common frame and epoch, e.g. ITRF2000, epoch 1997.0, we will be able to compare apples to apples. Remember that this change from epoch to epoch may be already included in the software that you are using. In any case, one should know that all coordinates must refer to a common epoch.

 

Finally, we are ready to transform between WGS84 (G1150) ≡ ITRF2000 at epoch 1997.0 and NAD83 (CORS96) at epoch 2002.0.

 

In order to do this final step, the 14 transformation parameters (three shifts, three rotations, one scale and their variations with time) between the two frames are needed. The rigorous transformation is somewhat involved [soler and Snay, 2004], thus, the easiest approach is to use the software developed at NGS which is interactively available through the Internet. Its name is Horizontal Time Dependant Positioning (HTDP) and can be located at the following URL: http://www.ngs.noaa.gov/TOOLS/Htdp/Htdp.shtml.

 

In fact, HTDP [snay, 1999] has also the capability of transforming coordinates from epoch to epoch using in the process model NNR-NUVEL-1A mentioned above. This software also could provide transformations from the WGS84 (G1150), assumed identical to ITRF2000, to NAD83 (CORS96). The text included on the HTDP Web page is self explanatory and sufficiently detailed for doing all types of transformations necessary when performing GPS work.

 

Final comment. Although on November 5, 2006, the GPS satellite orbits provided by the International GNSS Service switch to frame ITRF2005, everything mentioned above is still valid. Nothing practically will change considering that the differences between ITRF2000 and ITRF2005 are smaller than the differences between WGS84 (G1150) and ITRF2000. Except for accurate geodetic investigations these differences could be assumed negligible.

 

--------------------------------------------------------------------------------

 

References

1. Snay, R.A. (1999). Using HTDP software to transform spatial coordinates across time and between reference frames, Surveying and Land Information Systems, 59(1), 15-25.

http://www.ngs.noaa.gov/CORS/Articles/Using_HTDP.pdf

 

2. Soler, T. & R.A. Snay (2004). Transforming positions and velocities between the International Terrestrial Reference Frame of 2000 and North American Datum of 1983, J. Surv. Engrg., ASCE, 130(2), 49-55. http://www.ngs.noaa.gov/CORS/Articles/SolerSnayASCE.pdf

 

3. True, S.A. (2004) “Planning the future of the World Geodetic System 1984.” Proc. IEEE Position Location and Navigation Symposium, Monterey, CA, 26-29 April 2004, 10 p.

 

Further Reading

Below is a list of references that expand into the theory and practical applications of a topic as important as frame transformations for GPS users. We suggest that the persons desiring more information should start by reading and comprehend the series of articles about coordinate frames and datums written by Snay and Soler. For those interested on expanding into the theoretical intricacies of the problem a more technical set of citations are included. The reason that we are mainly concentrating on NGS publications is because all these references can be downloaded for free from the CORS Web page.

 

1. Snay, R.A. & T. Soler (1999). Part 1 - Modern Terrestrial Reference Systems. Professional Surveyor, 19(10), 32-33.

 

2. http://www.ngs.noaa.gov/CORS/Articles/Refe...tems-Part-1.pdf

 

3. Snay, R.A. & T. Soler (2000). Part 2 - The evolution of NAD83. Professional Surveyor, 20(2), 16, 18.

 

4. http://www.ngs.noaa.gov/CORS/Articles/Refe...tems-Part-2.pdf

 

5. Snay, R.A. & T. Soler (2000). Part 3 - WGS 84 and ITRS. Professional Surveyor, 20(3), 24, 26, 28.

 

6. http://www.ngs.noaa.gov/CORS/Articles/Refe...tems-Part-3.pdf

 

7. Snay, R.A. & T. Soler (2000). Part 4 - Practical considerations for accurate positioning. Professional Surveyor, 20(4), 32-34.

 

8. http://www.ngs.noaa.gov/CORS/Articles/Refe...tems-Part-4.pdf

 

9. Snay, R.A. (2003). Horizontal Time-Dependent Positioning, Professional Surveyor, 23(11), 30, 32, 34. ( http://www.ngs.noaa.gov/CORS/Articles/HTDPSnayPS.pdf )

 

10. Snay, R.A. (2003). Introducing two spatial reference frames for regions of the Pacific Ocean, Surveying and Land Information Science, 63(1), 5-12. http://www.ngs.noaa.gov/CORS/Articles/SalisSnay.pdf

 

11. Soler, T. & J. Marshall. (2003). A note on frame transformations with applications to geodetic datums, GPS Solutions, 7(1), 23-32. http://www.ngs.noaa.gov/CORS/Articles/GPSTrans2.pdf

 

12. Soler, T. & J. Marshall. (2002). Rigorous transformation of variance-covariance matrices of GPS-derived coordinates and velocities, GPS Solutions, 6(1-2), 76-90. http://www.ngs.noaa.gov/CORS/Articles/GPSTrans1.pdf

 

13. Soler, T., N.D. Weston, & H. Han. (2002). Computing NAD 83 coordinates using ITRF-derived vector components, Proc. XXII FIG International Congress, ACSM/ASPR Annual Conference, April 19-26, Washington D.C., 6 pg. http://www.ngs.noaa.gov/CORS/Articles/FIGSoler.doc

 

14. Soler, T. (2001). Densifying 3D GPS networks by accurate transformation of vector components, GPS Solutions, 4(3), 27-33. http://www.ngs.noaa.gov/CORS/Articles/Densifying3D.PDF

 

15. Soler, T. (1998). A compendium of transformation formulas useful in GPS work, Journal of Geodesy, 72(7-8), 482-490. ( http://www.ngs.noaa.gov/CORS/Articles/gpsPDF.pdf )

 

 

http://forums.gpsworld.com/showthread.php?t=17

[+NightPilot]

All that may be of some academic interest to some, but the practical answer for geocachers is that NAD83 and WGS84 are identical. Any differences are much smaller than the accuracy of the GPS, so they can be, for all practical purposes, considered to be the same.

[+Kewaneh & Shark]

The differences between the two datums are often discussed in the benchmark forums. This thread titled 'Nad 83 - Wgs 84' was started in May of 2005 by DaveD. It has some good discussion about the subject. It should be said here that Dave is the Chief Geodetic Surveyor for the National Geodetic Survey and he had a large hand in the development of the NAD83 datum.

 

This thread also has some good info.

 

- Kewaneh